The present disclosure relates generally to updating geospatial data, and more specifically to updating an old image's vector data to fit a new image of the same geographic area.
Geospatial data is information that identifies a geographic location and characteristics of natural or constructed features and boundaries on the earth. In some known systems, geospatial data location is sometimes collected from images of the location. The images may be captured using visible light cameras, near infrared (NIR) sensors, radar, thermal imaging sensors, or any other suitable sensor for capturing images of a geographic location. The images are often captured by sensors in one or more satellites or aerial vehicles. In some known systems, the images are analyzed to identify features shown in the images using vectors. The vectors are typically polygons, lines (or arcs), or point data that are used to represent attributes shown in the image being analyzed. Polygons are typically used to represent boundaries of areas, buildings, lakes, and the like. Lines (or arcs) are typically used to represent relatively linear features, such as roads, rivers, trails, etc. Point vector data is commonly used to identify discrete data points of interest in the image. Collecting vector data from images is generally a relatively slow and costly manual process. Vector data generated from images typically will not line up with the newer images of the same location due to differences in sensor, height, angle of the image, etc. Thus, when newer images of a location are acquired, all vector data for the location must be recollected (even vector data for unchanged features). It would be beneficial to update old vector data to new images automatically or semi-automatically.